Seixas, J, Carvalhais, N, Nunes, C, Benali, A (2009). Comparative analysis of MODIS-FAPAR and MERIS-MGVI datasets: Potential impacts on ecosystem modeling. REMOTE SENSING OF ENVIRONMENT, 113(12), 2547-2559.
Assessing the ecosystem productivity dynamics is a key issue in the context of the carbon cycle at the southern Europe ecosystems. The availability of remote sensing data and products, as FAPAR. strongly motivates its use in ecosystem modeling. This paper presents a comparative analysis between MODIS-FAPAR and MERIS-MGVI datasets, two products carrying similar information, from regional (the Iberian Peninsula) to local (six biomes) scales. Moreover, an assessment of the impacts of its use in terrestrial ecosystem modeling, by using the CASA production efficiency model is presented. The analysis was conducted for both datasets for the year 2003, and was supported by an extensive statistical analysis, including geostatistics, focusing both seasonality and spatial coherence. Main findings refer to (i) a systematically higher MODIS than MERIS data. both at the Iberian Peninsula and at local biomes, (ii) a seasonality of the correlation between the two datasets, and (iii) a spatial homogeneity over larger extensions from MERIS-MGVI data, opposing to a higher spatial variability from MODIS-FAPAR. Although these different data properties, it was found that the two independent datasets produce reasonable NEP estimates, by using CASA model at the site level. These results suggest that the MODIS-FAPAR and MERIS-MGVI differences can be minimized in biophysical modeling through model re-parameterization, although this process does not solve the seasonal dependence pattern found in the original datasets. (C) 2009 Elsevier Inc. All rights reserved.